-
Notifications
You must be signed in to change notification settings - Fork 0
Page Responsibilities
awagener edited this page Dec 17, 2012
·
2 revisions
This is both hilariously incomplete and hilariously illogical to anyone who isn't me. But perhaps it'll be of interest.
ATTRIBUTES
attr_accessible :title
attr_accessible :id, :deletable, :link_url, :menu_match, :meta_keywords,
:skip_to_first_child, :position, :show_in_menu, :draft,
:parts_attributes, :browser_title, :meta_description,
:parent_id, :menu_title, :page_id, :layout_template,
:view_template, :custom_slug, :slug
attr_accessor :locale # to hold temporarily
has_many :parts,
:foreign_key => :refinery_page_id,
:class_name => '::Refinery::PagePart',
:order => 'position ASC',
:inverse_of => :page,
:dependent => :destroy,
:include => ((:translations) if ::Refinery::PagePart.respond_to?(:translation_class))
accepts_nested_attributes_for :parts, :allow_destroy => true
----------
* Page data
attr_accessible :title
validates :title, :presence => true
before_destroy :deletable?
after_save :reposition_parts!
# Returns in cascading order: custom_slug or menu_title or title depending on
# which attribute is first found to be present for this page.
def custom_slug_or_title
custom_slug.presence || menu_title.presence || title
end
# Returns true if this page is "published"
def live?
not draft?
end
# Returns true if this page is the home page or links to it.
def home?
link_url == '/'
end
# Accessor method to test whether a page part
# exists and has content for this page.
# Example:
#
# ::Refinery::Page.first.content_for?(:body)
#
# Will return true if the page has a body page part and it is not blank.
def content_for?(part_title)
content_for(part_title).present?
end
# Accessor method to get a page part from a page.
# Example:
#
# ::Refinery::Page.first.content_for(:body)
#
# Will return the body page part of the first page.
def content_for(part_title)
part_with_title(part_title).try(:body)
end
# Accessor method to get a page part object from a page.
# Example:
#
# ::Refinery::Page.first.part_with_title(:body)
#
# Will return the Refinery::PagePart object with that title using the first page.
def part_with_title(part_title)
# self.parts is usually already eager loaded so we can now just grab
# the first element matching the title we specified.
self.parts.detect do |part|
part.title.present? and # protecting against the problem that occurs when have nil title
part.title == part_title.to_s or
part.title.downcase.gsub(" ", "_") == part_title.to_s.downcase.gsub(" ", "_")
end
end
# Used to index all the content on this page so it can be easily searched.
def all_page_part_content
parts.map(&:body).join(" ")
end
# Live pages are 'allowed' to be shown in the frontend of your website.
# By default, this is all pages that are not set as 'draft'.
def self.live
where(:draft => false)
end
# Finds pages by their title. This method is necessary because pages
# are translated which means the title attribute does not exist on the
# pages table thus requiring us to find the attribute on the translations table
# and then join to the pages table again to return the associated record.
def self.by_title(title)
with_globalize(:title => title)
end
----------
* Search (acts_as_indexed)
require 'acts_as_indexed'
# Docs for acts_as_indexed http://github.com/dougal/acts_as_indexed
acts_as_indexed :fields => [:title, :meta_keywords, :meta_description,
:menu_title, :browser_title, :all_page_part_content]
----------
* Url (friendly_id)
require 'friendly_id'
extend FriendlyId
validates :custom_slug, :uniqueness => true, :allow_blank => true
# Docs for friendly_id http://github.com/norman/friendly_id
friendly_id :custom_slug_or_title, :use => [:reserved, :globalize, :scoped],
:reserved_words => %w(index new session login logout users refinery admin images wymiframe),
:scope => :parent
# When this page is rendered in the navigation, where should it link?
# If a custom "link_url" is set, it uses that otherwise it defaults to a normal page URL.
# The "link_url" is often used to link to a plugin rather than a page.
#
# For example if I had a "Contact" page I don't want it to just render a contact us page
# I want it to show the Inquiries form so I can collect inquiries. So I would set the "link_url"
# to "/contact"
def url
if link_url.present?
link_url_localised?
elsif Refinery::Pages.marketable_urls
with_locale_param url_marketable
elsif to_param.present?
with_locale_param url_normal
end
end
# Add 'marketable url' attributes into this page's url.
# This sets 'path' as the nested_url value and sets 'id' to nil.
# For example, this might evaluate to /about for the "About" page.
def url_marketable
# :id => nil is important to prevent any other params[:id] from interfering with this route.
url_normal.merge :path => nested_url, :id => nil
end
# Returns a url suitable to be used in url_for in Rails (such as link_to).
# For example, this might evaluate to /pages/about for the "About" page.
def url_normal
{:controller => '/refinery/pages', :action => 'show', :path => nil, :id => to_param, :only_path => true}
end
def uncached_nested_url
[parent.try(:uncached_nested_url), to_param.to_s].compact.flatten
end
# Returns an array with all ancestors to_param, allow with its own
# Ex: with an About page and a Mission underneath,
# ::Refinery::Page.find('mission').nested_url would return:
#
# ['about', 'mission']
#
alias_method :nested_url, :uncached_nested_url
# Returns the string version of nested_url, i.e., the path that should be generated
# by the router
def nested_path
['', nested_url].join('/')
end
##
# Protects generated slugs from title if they are in the list of reserved words
# This applies mostly to plugin-generated pages.
# This only kicks in when Refinery::Pages.marketable_urls is enabled.
#
# Returns the sluggified string
def normalize_friendly_id_with_marketable_urls(slug_string)
sluggified = slug_string.to_slug.normalize!
if Refinery::Pages.marketable_urls && self.class.friendly_id_config.reserved_words.include?(sluggified)
sluggified << "-page"
end
sluggified
end
alias_method_chain :normalize_friendly_id, :marketable_urls
# With slugs scoped to the parent page we need to find a page by its full path.
# For example with about/example we would need to find 'about' and then its child
# called 'example' otherwise it may clash with another page called /example.
def self.find_by_path(path)
split_path = path.to_s.split('/').reject(&:blank?)
page = ::Refinery::Page.by_slug(split_path.shift, :parent_id => nil).first
page = page.children.by_slug(split_path.shift).first until page.nil? || split_path.empty?
page
end
# Helps to resolve the situation where you have a path and an id
# and if the path is unfriendly then a different finder method is required
# than find_by_path.
def self.find_by_path_or_id(path, id)
if Refinery::Pages.marketable_urls && path.present?
if path.friendly_id?
find_by_path(path)
else
find(path)
end
elsif id.present?
find(id)
end
end
# Finds pages by their slug. See by_title
def self.by_slug(slug, conditions={})
locales = Refinery.i18n_enabled? ? Refinery::I18n.frontend_locales.map(&:to_s) : ::I18n.locale.to_s
with_globalize({ :locale => locales, :slug => slug }.merge(conditions))
end
----------
* ???
# when collecting the pages path how is each of the pages seperated?
PATH_SEPARATOR = " - "
# Used for the browser title to get the full path to this page
# It automatically prints out this page title and all of it's parent page titles joined by a PATH_SEPARATOR
def path(options = {})
# Override default options with any supplied.
options = {:reversed => true}.merge(options)
unless parent_id.nil?
parts = [title, parent.path(options)]
parts.reverse! if options[:reversed]
parts.join(PATH_SEPARATOR)
else
title
end
end
----------
* Translations
translates :title, :menu_title, :custom_slug, :slug, :include => :seo_meta
class Translation
is_seo_meta
attr_accessible :browser_title, :meta_description, :meta_keywords, :locale
end
# Delegate SEO Attributes to globalize3 translation
seo_fields = ::SeoMeta.attributes.keys.map{|a| [a, :"#{a}="]}.flatten
delegate(*(seo_fields << {:to => :translation}))
has_many :parts,
:foreign_key => :refinery_page_id,
:class_name => '::Refinery::PagePart',
:order => 'position ASC',
:inverse_of => :page,
:dependent => :destroy,
:include => ((:translations) if ::Refinery::PagePart.respond_to?(:translation_class))
before_save { |m| m.translation.save }
* What does 'page' mean in this sentence?
# The canonical page for this particular page.
# Consists of:
# * The default locale's translated slug
def canonical
Globalize.with_locale(::Refinery::I18n.default_frontend_locale){ url }
end
# The canonical slug for this particular page.
# This is the slug for the default frontend locale.
def canonical_slug
Globalize.with_locale(::Refinery::I18n.default_frontend_locale) { slug }
end
# Adds the locale key into the URI for this page's link_url attribute, unless
# the current locale is set as the default locale.
def link_url_localised?
return link_url unless ::Refinery.i18n_enabled?
current_url = link_url
if current_url =~ %r{^/} && ::Refinery::I18n.current_frontend_locale != ::Refinery::I18n.default_frontend_locale
current_url = "/#{::Refinery::I18n.current_frontend_locale}#{current_url}"
end
current_url
end
# If the current locale is set to something other than the default locale
# then the :locale attribute will be set on the url hash, otherwise it won't be.
def with_locale_param(url_hash, locale = nil)
locale ||= ::Refinery::I18n.current_frontend_locale if self.class.different_frontend_locale?
url_hash.update :locale => locale if locale
url_hash
end
# Wrap up the logic of finding the pages based on the translations table.
def self.with_globalize(conditions = {})
conditions = {:locale => ::Globalize.locale.to_s}.merge(conditions)
globalized_conditions = {}
conditions.keys.each do |key|
if (translated_attribute_names.map(&:to_s) | %w(locale)).include?(key.to_s)
globalized_conditions["#{self.translation_class.table_name}.#{key}"] = conditions.delete(key)
end
end
# A join implies readonly which we don't really want.
joins(:translations).where(globalized_conditions).where(conditions).readonly(false)
end
----------
* SEO
# Delegate SEO Attributes to globalize3 translation
seo_fields = ::SeoMeta.attributes.keys.map{|a| [a, :"#{a}="]}.flatten
delegate(*(seo_fields << {:to => :translation}))
----------
* International locale
before_create :ensure_locale, :if => proc { ::Refinery.i18n_enabled? }
# Make sures that a translation exists for this page.
# The translation is set to the default frontend locale.
def ensure_locale
if self.translations.empty?
self.translations.build(:locale => (::Refinery::I18n.default_frontend_locale if Refinery.i18n_enabled?))
end
end
private :ensure_locale
* This calls nothing local to this model
# Wraps up all the checks that we need to do to figure out whether
# the current frontend locale is different to the current one set by ::I18n.locale.
# This terminates in a false if i18n extension is not defined or enabled.
def self.different_frontend_locale?
::Refinery.i18n_enabled? && ::Refinery::I18n.current_frontend_locale != ::I18n.locale
end
----------
* Menu
# Docs for acts_as_nested_set https://github.com/collectiveidea/awesome_nested_set
# rather than :delete_all we want :destroy
acts_as_nested_set :dependent => :destroy
# Return true if this page can be shown in the navigation.
# If it's a draft or is set to not show in the menu it will return false.
def in_menu?
live? && show_in_menu?
end
def not_in_menu?
not in_menu?
end
# Returns all visible sibling pages that can be rendered for the menu
def shown_siblings
siblings.reject(&:not_in_menu?)
end
def refinery_menu_title
[menu_title, title].detect(&:present?)
end
def to_refinery_menu_item
{
:id => id,
:lft => lft,
:menu_match => menu_match,
:parent_id => parent_id,
:rgt => rgt,
:title => refinery_menu_title,
:type => self.class.name,
:url => url
}
end
# Shows all pages with :show_in_menu set to true, but it also
# rejects any page that has not been translated to the current locale.
# This works using a query against the translated content first and then
# using all of the page_ids we further filter against this model's table.
def self.in_menu
where(:show_in_menu => true).with_globalize
end
def self.fast_menu
live.in_menu.order('lft ASC').includes(:translations)
end
----------
* Caching
after_save :expire_page_caching
after_destroy :expire_page_caching
def expire_page_caching
self.class.expire_page_caching
end
private :expire_page_caching
def self.expire_page_caching
begin
Rails.cache.delete_matched(/.*pages.*/)
rescue NotImplementedError
Rails.cache.clear
warn "**** [REFINERY] The cache store you are using is not compatible with Rails.cache#delete_matched - clearing entire cache instead ***"
ensure
return true # so that other callbacks process.
end
end
----------
* Deletion
# Am I allowed to delete this page?
# If a link_url is set we don't want to break the link so we don't allow them to delete
# If deletable is set to false then we don't allow this page to be deleted. These are often Refinery system pages
def deletable?
deletable && link_url.blank? and menu_match.blank?
end
# Before destroying a page we check to see if it's a deletable page or not
# Refinery system pages are not deletable.
def destroy
return super if deletable?
unless Rails.env.test?
# give useful feedback when trying to delete from console
puts "This page is not deletable. Please use .destroy! if you really want it deleted "
puts "unset .link_url," if link_url.present?
puts "unset .menu_match," if menu_match.present?
puts "set .deletable to true" unless deletable
end
false
end
# If you want to destroy a page that is set to be not deletable this is the way to do it.
def destroy!
self.menu_match = nil
self.link_url = nil
self.deletable = true
destroy
end
----------
* Page Parts
# Repositions the child page_parts that belong to this page.
# This ensures that they are in the correct 0,1,2,3,4... etc order.
def reposition_parts!
reload.parts.each_with_index do |part, index|
part.update_attributes :position => index
end
end
----------
* Pagination
# Returns how many pages per page should there be when paginating pages
def self.per_page(dialog = false)
dialog ? Pages.pages_per_dialog : Pages.pages_per_admin_index
end
Menu item stuff should be its own model
URL stuff should be its own model
Is generalising out search and slug stuff a good or the most terriblest idea ever?